Serum 25-hydroxyvitamin D measurement: Comparative evaluation of three automated immunoassays

Objectives: the different analytical methods for measurement of serum 25-hydroxyvitamin D (25(OH)D) are not yet fully harmonized and no consensus exists on a threshold of 25(OH)D defining a deficiency status. In this study, we compared the results from the assays of serum 25(OH)D performed with three different methods to evaluate the presence of potential biases and how much these biases can influence the assignment of patients to specific 25(OH)D deficiency/sufficiency categories. Design and Methods: Liaison 25(OH) Vitamin D Total (DiaSorin Liaison XL), Elecsys Vitamin D Total II (Roche Elecsys) and Lumipulse G25(OH) Vitamin D (Fujirebio Lumipulse G1200) were used. Methods comparability was established performing Passing-Bablok regression and Bland-Altman analysis to prove whether the differences found were lower than the preliminarily pre-established maximum acceptable bias. Results: all Passing-Bablok regressions exhibited the presence of a proportional and constant systematic error. Bland-Altman analysis revealed biases well above the maximum acceptable bias, so the 25(OH)D concentrations measured were not comparable. To evaluate whether the three methods had the same ability to classify patients into different categories of vitamin D levels, we categorized results obtained by each method in reference classes. Lumipulse categorized most patients into the class with the lowest 25(OH)D concentrations (